Find out how good/bad your battery is

[e-gadget-guy]

interesting battery log... only included three lines before and after change from centro 1150 to the 2250...

2010-08-12T04:54:09.152435Z [3160] palm-webos-device user.info powerd: {powerd}: BATTERY_IPC: (841.599976mAh, 87%, 91%_ui, 34C, 68mA, 4201mV, authentic)
2010-08-12T05:01:02.989044Z [3574] palm-webos-device user.info powerd: {powerd}: BATTERY_IPC: (848.000000mAh, 87%, 91%_ui, 33C, 42mA, 4196mV, authentic)
2010-08-12T05:01:23.672332Z [3595] palm-webos-device user.info powerd: {powerd}: CHG_LOGIC: debug_battery_taper (charging, type=0, 848.000000mAh, 87%, 33C, 42mA, 4196mV, (avg)40mA, min=50mA, max=4141mV, taper_state=start, battery_state=undervoltage, battery-authentic charger-connected)
2010-08-12T05:01:53.625396Z [3625] palm-webos-device user.warning powerd: {powerd}: w1_battery: Bogus battery level 100%
2010-08-12T05:01:53.635589Z [3625] palm-webos-device user.warning powerd: {powerd}: w1_battery: Bogus battery level 100%
2010-08-12T05:01:53.646026Z [3625] palm-webos-device user.warning powerd: {powerd}: w1_battery: Bogus battery level 100%
2010-08-12T05:01:53.656616Z [3625] palm-webos-device user.warning powerd: {powerd}: w1_battery: Bogus battery level 100%
2010-08-12T05:01:53.666900Z [3625] palm-webos-device user.warning powerd: {powerd}: w1_battery: Bogus battery level 100%
2010-08-12T05:01:53.677947Z [3625] palm-webos-device user.warning powerd: {powerd}: w1_battery: Bogus battery level 100%
2010-08-12T05:01:53.745727Z [3625] palm-webos-device user.info powerd: {powerd}: BATTERY_IPC: (969.599976mAh, 100%, 100%_ui, 33C, 44mA, 4196mV, authentic)
2010-08-12T05:01:53.767639Z [3625] palm-webos-device user.info powerd: {powerd}: CHG_LOGIC: debug_battery_taper (charging, type=0, 969.599976mAh, 100%, 33C, 44mA, 4196mV, (avg)43mA, min=50mA, max=4141mV, taper_state=complete, battery_state=undervoltage,charge-termination, battery-authentic charger-connected)
2010-08-12T05:00:57.122528Z [59] palm-webos-device user.info powerd: {powerd}: BATTERYPOLL: BatteryState debounce
2010-08-12T05:00:57.303192Z [59] palm-webos-device user.info powerd: {powerd}: BATTERYPOLL: BatteryState inserted
2010-08-12T05:00:57.464630Z [59] palm-webos-device user.info powerd: {powerd}: BATTERYPOLL: BatteryState authentic
2010-08-12T05:00:57.667266Z [59] palm-webos-device user.info powerd: {powerd}: BATTERY_IPC: (2081.600098mAh, 96%, 100%_ui, 31C, -461mA, 4074mV, authentic)
2010-08-12T05:00:58.730987Z [61] palm-webos-device user.info powerd: {powerd}: BATTERY_IPC: (2081.600098mAh, 96%, 100%_ui, 31C, -461mA, 4074mV, authentic)
2010-08-12T05:06:06.186279Z [96] palm-webos-device user.info powerd: {powerd}: BATTERY_IPC: (2076.800049mAh, 96%, 100%_ui, 32C, -821mA, 4045mV, authentic)

[e-gadget-guy]

Originally Posted by e-gadget-guy

interesting battery log... only included three lines before and after change from centro 1150 to the 2250...
snip
2010-08-12T05:06:06.186279Z [96] palm-webos-device user.info powerd: {powerd}: BATTERY_IPC: (2076.800049mAh, 96%, 100%_ui, 32C, -821mA, 4045mV, authentic)

also, it stopped at 100% showing 2156.800049mAh and no more bogus messages.

[somline]

Originally Posted by e-gadget-guy

BATTERY_IPC: (848.000000mAh, 87%, 91%_ui, 33C, 42mA, 4196mV, authentic)
powerd: {powerd}: w1_battery: Bogus battery level 100%
BATTERY_IPC: (969.599976mAh, 100%, 100%_ui, 33C, 44mA, 4196mV, authentic)

This is what we find when the battery is lying:
In this case a jump from 87% to 100%. This battery has actually only around 850mAh and not 969mAh as (false) reported.

Originally Posted by e-gadget-guy

2010-08-12T05:06:06.186279Z [96] palm-webos-device user.info powerd: {powerd}: BATTERY_IPC: (2076.800049mAh, 96%, 100%_ui, 32C, -821mA, 4045mV, authentic)

This battery seems to have around 2100mAh maybe more. Find the 100% and look at the entry before.
EDIT: Mist your second post: Good battery ;-)

Question: Did you shutdown the device for exchanging the batteries?

[Xanadu73]

somline,

Unless I'm missing something, I don't see how to attach text files to a PM, so I'm doing this here.

Here's my (grep -i battery /var/log/messages) from last night (your morning!). It clearly shows that it drops from 17% --> 0%. It complains about "bogus battery level 0%". You then see when I plugged in my cable. Then it charges to 82% and then it gets weird... :-(

battery-08-11-2010.txt



M.

[somline]

Originally Posted by Xanadu73

Here's my (grep -i battery /var/log/messages) from last night (your morning!). It clearly shows that it drops from 17% --> 0%. It complains about "bogus battery level 0%". You then see when I plugged in my cable. Then it charges to 82% and then it gets weird... :-(

Hi Xanadu73,
some good and some not so good news:

1. You got the battery into lerningmode ;-)
2. The pre decided to stop the learning because the voltage has reached the maximum of 4191mV

I will investigate your register settings when I get time. Maybe the max voltage settings in your battery have changed for some reason.

In the meantime we can let the battery "learn" by hand:
Your battery has a capacity of around 1100mAh at 82%.
(Last BATTERY_IPC entry before 100%)
So you can set the age to say 85.15625% with this command:

cd /sys/devices/w1*/32-*
echo 0x146d > setreg

This should repair the sudden shutdown.

EDIT: If you want to play with other values:
ATTENTION: Don't mistype the 0x14!! You can do real damage to the battery/pre
0x14 is the address of the register for the age value
80 is 100% and 40 is 50%
0x1480 = 100%
0x146d = 85,15...%
0x1440 = 50%

EDIT2: After changing this value the pre may get a litle confused. After next discharge/charge it should be fine.

[Xanadu73]

Originally Posted by somline

Hi Xanadu73,
some good and some not so good news:

1. You got the battery into lerningmode ;-)
2. The pre decided to stop the learning because the voltage has reached the maximum of 4191mV

I will investigate your register settings when I get time. Maybe the max voltage settings in your battery have changed for some reason.

In the meantime we can let the battery "learn" by hand:
Your battery has a capacity of around 1100mAh at 82%.
(Last BATTERY_IPC entry before 100%)
So you can set the age to say 85.15625% with this command:

cd /sys/devices/w1*/32-*
echo 0x146d > setreg

This should repair the sudden shutdown.

EDIT: If you want to play with other values:
ATTENTION: Don't mistype the 0x14!! You can do real damage to the battery/pre
0x14 is the address of the register for the age value
80 is 100% and 40 is 50%
0x1480 = 100%
0x146d = 85,15...%
0x1440 = 50%

EDIT2: After changing this value the pre may get a little confused. After next discharge/charge it should be fine.

Interesting.

Here are my findings.

0x146d - Make the battery think it only holds the ~1100 you say it will.
0x1480 - Makes the battery think it only holds ~9mAh LOL!

Cool part:
After playing with different values, I found that 0x147f makes the battery think it hold the full 1333mAh!






I 100% understand I'm playing in dangerous territory here and that you in no way, shape or form are responsible for me melting my battery and or phone since you have gone on record as stating that you do not recommend this.

To anyone else reading these words, you need to understand exactly what I just typed. You could loose everything and Palm & your carrier is not going to care.



M.

[somline]

Originally Posted by Xanadu73

Interesting.

Here are my findings.

0x146d - Make the battery think it only holds the ~1100 you say it will.
0x1480 - Makes the battery think it only holds ~9mAh LOL!

Cool part:
After playing with different values, I found that 0x147f makes the battery think it hold the full 1333mAh!

Yes indeed interesting/strange,
I'm afraid it will not solve your sudden death problem. But if you have the time why not give it a shot.
Keep me informed

[somline]

For those knowing what a shell is (other than an oil company) and want to recalibrate the battery I've build a helper script.
This script will NOT change anything in your palm or battery.
If the calibration works the battery chip will change the AGE to an actual value.

The calibration begins when the battery has reached very low voltage (3,4V).
The script will inform you with a text and sound because now you have to connect the charger immediately. You don't have much time before the pre will shutdown.
Then the script monitors the calibration. Don't interrupt the charging now as this will stop the calibration.
When the calibration is finished you get a message.

Unfortunately sometimes the pre may stop the calibration if it thinks the battery is full. This will also stop calibrating. Xanadu73's post is an example for this. For some reasons sometimes the calibration will be interrupted by connecting the charger.
In all this cases you get an error message from the script.

I may build an ipk later if this script works and is useful (certainly for free).
So please try it and report your results.

Start the script as root with:

sh <path>/bcm.sh
or make it executabe with  
chmod 755 <path>/bcm.sh and then call it with <path>/bcm.sh

EDIT: you can interrupt the script with <ctrl>c (sym button c in terminal) at anytime and restart at any time.
If it doesn't stop open another shell and: killall sh

EDIT2: Use nodoze to keep pre from sleeping
Thanks

[jp99]

Originally Posted by somline

The charging (and discharging) can be found by searching for "BATTERY_IPC":

{powerd}: BATTERY_IPC: (22.400000mAh, 2%, 2%_ui, 32C, 128mA, 3830mV, authentic)

The first value (22.400000mAh in this case) is showing the percentage of the battery.
so from the top find the first 100% and go bak up to the first BATTERY_IPC entry before that 100%. Check the timestamp to find the correct part.
Look for every line containing the word "BATTERY" or "battery" between those places(the 100% and the % before 100) like:
w1_battery: Bogus battery level 100%
CHG_LOGIC: BatteryOverchargeFault overcharge diag:

Look at davsalinas post for an example.

You should do something better on vacation ;-)

I am enjoying vacation - mostly hiking in mountains. I have a few rest days and like keeping up with the Pre world while I'm out. Even 1.4.5 came out while I was away. I have no idea what else is going on in the world though!

I have a number of message files (most compressed) but strangely no BATTERY_IPC lines in them (messages, and messages.<0 to 4>.gz are the files). I also searched "battery" but no charging lines appeared.

I'm thinking I have some switch set to less logging? This Pre is a new refurb to me (about 3 weeks old); I don't recall if I changed the logging parameters when I set it up.

[probowler2009]

Ok went to the Sprint Store to see if they could replace my battery, and they told me they will need to run a test and it would take 3 hours to run. They gave me a used replacement battery to use while they tested mine. Went back to the store to find out that the battery passed the test and they returned it to me. Put it back in my phone and got these results:

getfull40 - 1183.125
getage - 100.00
getcapacity 1182.980

Not sure if what they tested the battery on cleared out the charging memory. The battery holds the charge really well. I say visit your Sprint Store and have them run a test on it and it may revive your battery.

[davsalinas]

Finally got my Seidio 1350mAh battery into learning mode, using somline helper script to .... well, help. Took me two tries because on the first try, I got some sort of error, very near the end of the charge cycle, so I was not confident that it worked properly. Today, I tried again and this time it went all the way through. Here are the results.

2010-08-12T13:36:11.253387Z : BATTERY_IPC: (241.600006mAh, 19%, 20%_ui, 30C, -147mA, 3445mV, authentic)
2010-08-12T13:37:31.193908Z : w1_battery: Bogus battery level 0%
2010-08-12T13:37:31.247772Z : w1_battery: Bogus battery level 0%
2010-08-12T13:37:31.321197Z : BATTERY_IPC: (0.000000mAh, 0%, 0%_ui, 30C, -185mA, 3411mV, authentic)
2010-08-12T13:37:39.147674Z : CHG_LOGIC: ChargeStateTransitionLog in charging (P: 0%, T: 30 C, C: -218 mA, V: 3425 mV, AUTH true, Charger enabled: usb @ 1000mA)
2010-08-12T13:38:19.292053Z : BATTERY_IPC: (4.800000mAh, 0%, 0%_ui, 32C, 583mA, 3596mV, authentic)
2010-08-12T13:38:44.252471Z : BATTERY_IPC: (9.600000mAh, 1%, 1%_ui, 32C, 570mA, 3625mV, authentic)
........... deleted a lots of stuff here ...............
2010-08-12T15:54:23.609832Z : BATTERY_IPC: (1016.000000mAh, 82%, 86%_ui, 31C, 76mA, 4191mV, authentic)
2010-08-12T16:04:53.550231Z : w1_battery: Bogus battery level 100%
2010-08-12T16:04:53.601776Z : w1_battery: Bogus battery level 100%
2010-08-12T16:04:53.667297Z : BATTERY_IPC: (1024.000000mAh, 100%, 100%_ui, 30C, 56mA, 4201mV, authentic)
2010-08-12T16:09:33.612365Z : CHG_LOGIC: debug_battery_taper (charging, type=0, 1027.199951mAh, 100%, 30C, 44mA, 4196mV, (avg)48mA, min=50mA, max=4141mV, taper_state=start, battery_state=power-on-reset,charge-termination, battery-authentic charger-connected)
2010-08-12T16:10:03.620758Z : CHG_LOGIC: debug_battery_taper (charging, type=0, 1028.800049mAh, 100%, 30C, 48mA, 4201mV, (avg)48mA, min=50mA, max=4141mV, taper_state=complete, battery_state=power-on-reset,charge-termination, battery-authentic charger-connected)
2010-08-12T16:10:03.621154Z : CHG_LOGIC: Turning charging off because of charge complete
2010-08-12T16:10:03.621398Z : CHG_LOGIC: ChargeStateTransitionLog in chargecomplete (P: 100%, T: 30 C, C: 48 mA, V: 4201 mV, AUTH true, Charger disabled: usb @ 1000mA)
2010-08-12T16:10:03.733673Z : CHG_LOGIC: Sucessful chargeSourceSet none @ 0.
2010-08-12T16:10:03.734100Z : CHG_LOGIC: ChargingLogicChargerEnabled: Off charging + wall charger, turn bypass ON
2010-08-12T16:21:58.608642Z : BATTERY_IPC: (1017.599976mAh, 99%, 100%_ui, 29C, -40mA, 4167mV, authentic)

By the way, I trimmed the log entries for easier readability, so your logs won't look exactly like this. I've seen that others with Seidio 1350 batteries were lucky enough to get proper capacity in the 1300 range, but it looks like mine was probably 1150 all along, which was my original getfull40 setting before I hard coded it to 1350 to see what that would do.

[somline]

Thanks davsalinas and sorry to hear that.
You hopefully get it replaced.
Can you please show us getfull40 and getage? Did the age changed by recalibrating?

[blackfireball5]

Just wanted to thank everyone involved in this thread. I hated my battery life and thought it was a result of the SR-71 overclock. Luckily, my getage was 62%.

Sprint tested my phone and i have a new battery coming in tomorrow covered under the insurance for sprint.

This thread deserves a spot on the main news page for precentral.net

[davsalinas]

Sure,
getfull40 = 1350 (learning mode did not change this)
getcapacity = 1042.546 (this dropped quite a bit)
getage = 77.34375 (this dropped quite a bit)

I don't remember exactly what getage was before, sorry, but I believe it was around 93% or so. After getting the battery/charger into learning mode, it looks like it finally recalculated the age, but it based that on the 1350 that I had previously updated myself, from the original 1150. Previously, I had noticed that when I bumped up the getfull40 from 1150 to 1350, getage stayed the same, and getcapacity jumped up over 1200, which makes sense since I had fiddled with the baseline of 1150 and made it 1350. Getting the battery into learning mode seems to have forced it to validate the actual current capacity of the battery and readjust the age, but still trusting the getfull40 value.

I'll probably reset getfull40 back to 1150 this weekend and force another battery learning mode session to see what happens, but I'm pretty sure it will determine the same capacity, and reset the getage back close to 90% as well since the baseline will be lower.

[somline]

Thanks, thats another proof of my theory and exactly what I expected.
Recalibrating will change getage (and therefore getcapacity) only.

And again: getfull40 is set by the manufactor and will never get changed by the pre or the batterychip. So seidio is selling 1350mAh batteries which are 1150.
I wrote another mail to seidio but this time I don't got any answer. I will give them one more day before posting a big warning about buying these batteries.

[jbg7474]

Wow! I'm stunned. Thanks for the great work!

[somline]

I've updated the first page to summarize all findings we made. Please check that there is nothing missing.
I want to thank all of you participating and for the confidence you have placed in me.
Especially davsalinas and Xanadu73.
So much fun to play with just a phone ;-)

[themirthfulswami]

Hey somline, is it possible that altering the getfull40 value from the original 1150 value to the 1350 (for Seidio extended battery) value could accelerate the aging of the battery? I've been keeping a very watchful eye on the various numbers since changing the settings in my Pre and my getage value dropped from 90.625 to 89.8734 yesterday. Could be coincidence I suppose

[somline]

Originally Posted by themirthfulswami

Hey somline, is it possible that altering the getfull40 value from the original 1150 value to the 1350 (for Seidio extended battery) value could accelerate the aging of the battery? I've been keeping a very watchful eye on the various numbers since changing the settings in my Pre and my getage value dropped from 90.625 to 89.8734 yesterday. Could be coincidence I suppose

Yes it's a coincidence. Getage gets reduced depending on how often you charge the battery.
From the manual:

"Aging Estimation
As discussed above, the AS register value is adjusted occasionally based on cumulative discharge. As the ACR register decrements during each discharge cycle, an internal counter is incremented until equal to 32 times the AC. The AS is then decremented by one, resulting in a decrease of the scaled full battery capacity by 0.78% (approximately 2.4% per 100 cycles). See the AC register description above for recommendations on customizing the age-estimation rate."
EDIT2: From 90.625 to 89.8734 is almost 0.78%

EDIT: Altering will not accelerate the aging but if the battery gets into learnmode (what can happen if you connect the charger at the right time) age will be completely recalculated and may drop to say 80% to reflect the fact that the battery doesn't have 1350 but 1150 (if it is a fake one).
This is what davsalinas got (but he re-calibrated intentionally)

[Banger]

1114.0625 is mine, i should try my other battery?